Public health surveillance programs often compare the geographic clustering of sexually transmitted diseases (STDs) with demographic and social features of neighborhoods to determine STD risk factors. The findings shape STD prevention and control strategies. Our research suggests that these strategies are important but may have limited success because they infer individual-level risk factors for STDs from associations between residential neighborhood demographics and rates of disease. Many adolescents and young adults who should be at high risk for STDs on the basis of the demographic composition and social ecology of their residential neighborhood are not infected, while other individuals who should be at low risk for STDs on the basis of the demographic composition and social ecology of their residential neighborhood, are infected. A goal of the proposed study is to address shortcomings of past ecologic research. The specific aims are: (1) to determine whether sex partner selection patterns, within and between neighborhoods, are associated with adolescents' and young adults' risk for STDs in Baltimore and whether these patterns change over time; (2) to determine whether the collective efficacy and social capital of neighborhoods are associated with the prevalence of disease in part due to the availability of high STD risk sex partners; and (3) to determine whether age differences in normative social behavior account for the extent to which individuals do or do not have sex with someone who engages in similar risk behaviors (assortative and disassortative sexual mixing, respectively). The proposed study will conduct interviews and collect urine specimens for gonorrhea and chlamydia infection testing among a household sample of 960 sexually experienced adolescents and young adults residing in core (high STD prevalence) and non-core (low STD prevalence) census block groups within Baltimore City. Interviews will be conducted in the respondent's household at baseline and one year later using audio computer-assisted self-interviewing. Urine specimens will be tested using ligase chain reaction. We will also interview locatable sex partners. In addition, we will conduct a qualitative sub-study, which will provide formative data for survey development and explanatory data to supplement survey findings.